We can get pervasive patient and environment information.

We can:
- Increase mobility
- Reduce Invasivity
- Improve data visualization
- Improve delirium recognition
- Bridge the gap between ICU and wards

#ESICMInnovation

AI will probably enter on different levels with different roles.

The current model not well-suited if we frame healthcare (and ICU)
as a complex societal function to optimize health and well-being.

AI 3.0 is more than a product and it could probably disrupt the entire structure of healthcare.

Con:
- to understand how the machine works
- to integrate it with our workflow
- Accuracy of the input data
- Ethical and legal implications

Hence we need
- Define and prioritize specific features /protocols to be automated
- Deep learning and understanding physiological structure

#ESICMInnovation

Hence, most of the challenges are upstream of the data, and not enough hospitals are testing their predictive AI model for accuracy or data.

So… rather than thinking about new models, think how will humans change the way they practice due to the AI input.

Food for thought!

Again, excellent speakers and faculty lead an exceptional and stimulating discussion at the end of each presentation!

Keep up the good work people!

4/4

#ESICMInnovation

Next, Ari Ercole discussed the feasibility of AI implementation in ICU.

The coolest, newest device is useless if it does not change the clinical practice.

Hence, we should focus on how to change our systems implementing AI for the benefit of our patients and not someone else.

3/

While we could debate all night about how AI has its pros and cons, we should rather focus on the role of ESICM and other societies in developing adequate guidelines and regulations, benefiting humanity and limit damage.

2/

I highly appreciate the style of the final discussions, that leaves a lot of space for the audience to interact with the speakers. We are hearing more about everyday aspects and reflecting about the real life impact of the decision that are still to be made on how to introduce AI in our field.

4/4

We will probably get to ULTRArapid diagnostic, precision medicine and novel therapeutics allowing for better targeting and medicine release. The final step will probably be AI driven CCSS.

3/

AI will probably first reduce the physical workload, while boosting access to ICU and it will then augment ICU MD’s expert judgement. It will replace some jobs while accelerating renewal.

Regarding therapies, one good example comes from antibiotics…

2/